def inciso3v(self):
codigo = int(
input("Ingrese codigo del producto que desea cambiar stock: "))
cantid = int(input("Ingrese la cantidad: "))
stock1 = Stock(codigo, cantid)
for i in range(len(self.__listaStock)):
if self.__listaStock[i].getCodigoProducto() == codigo:
stock2 = Stock(codigo, self.__listaStock[i].getCantidad())
restar = stock1 - stock2
self.inciso3a(codigo, restar)
def main():
chef = LineChef('BTC-USD', 20, 8, [1, 2, 3, 4])
startDate = datetime.datetime(2018, 2, 4, 8)
endDate = startDate + datetime.timedelta(days=60)
chef.train(startDate, endDate)
strongCount = 0
for pat in chef.patterns:
if (pat.isStrong()):
strongCount += 1
print("# Matches: ", pat.numMatches)
print(pat.changes)
print(pat)
print("STRONG COUNT: ", strongCount)
testStartDate = endDate
testEndDate = testStartDate + datetime.timedelta(days=14)
predictionDelta = 4
simBoi = RepeatPredictor(5, Stock('BTC-USD'), 1000, 20, testStartDate,
datetime.timedelta(hours=predictionDelta))
chef.test(testStartDate, testEndDate, predictionDelta, simBoi)
def loadValuesIntoOneStock(self):
testStock = Stock('TEST', 'testValues.txt')
testStock.record = {
"2019-03-18": {
"06:00:00": (0.5, 45.10),
"06:15:00": (73, 12)
},
"2019-03-19": {
"06:15:00": (83, 12)
},
"2019-03-20": {
"06:30:00": (14, -6.0),
"06:31:00": (18, 9)
}
}
return testStock
def manejadorArchiStock(self):
archi = open("Stock.csv")
reader = csv.reader(archi, delimiter=';')
next(reader)
for i in reader:
codP = int(i[0])
cant = int(i[1])
objeto = Stock(codP, cant)
self.agregadorStock(objeto)
archi.close()
class TestStock(TestCase):
def setUp(self):
self.TEA = Stock("TEA", "Common", last_dividend=0, par_value=100)
self.POP = Stock("POP", "Common", last_dividend=8, par_value=100)
self.ALE = Stock("ALE", "Common", last_dividend=23, par_value=60)
self.GIN_PRE = Stock("GIN", "Preferred", last_dividend=8, par_value=100, fixed_dividend=0.02)
self.GIN_COM = Stock("GIN", "Common", last_dividend=8, par_value=100)
# if last_dividend 0, we expect a dividend yield of 0.
def test_dividend_yield_1(self):
self.assertEqual(self.TEA.dividend_yield(35), 0)
def test_dividend_yield_2(self):
self.assertEqual(self.POP.dividend_yield(55), 8/55)
def test_dividend_yield_3(self):
self.assertEqual(self.GIN_PRE.dividend_yield(89), (0.02*100)/89)
def test_dividend_yield_4(self):
self.assertEqual(self.GIN_COM.dividend_yield(89), 8/89)
# if last dividend of a stock is 0, we set its P/E ratio to 0.
def test_p_e_ratio_1(self):
self.assertEqual(self.TEA.p_e_ratio(50), 0)
def test_p_e_ratio_2(self):
self.assertEqual(self.GIN_PRE.p_e_ratio(92), 92/8)
def main(transaction_history):
# this method assumes the transaction history is sent in the form of a dictionary.
# It processes the transaction history of each instrument and creates a portfolio based on it.
for key, value in transaction_history.iteritems():
for all_bonds, tran_history in value.iteritems():
transaction = get_current_quantity(tran_history)
if key == 'Stock':
instrument = Stock(all_bonds, transaction)
elif key == 'Bonds':
instrument = Bonds(all_bonds, transaction)
elif key == 'Cash':
instrument = Cash(all_bonds, transaction)
else:
raise ValueError('Instrument not defined')
all_instrument[all_bonds] = instrument
return create_portfolio()
def main():
# ts = TimeSeries(key=API_KEY)
# for key, stock in ALL_STOCKS.items():
# data, metaData = ts.get_intraday(symbol=stock.symbol, interval="5min", outputsize="full")
# for dataKey, item in data.items():
# stock.addValue(dataKey.split(" ")[1], item["4. close"], item["5. volume"], dataKey.split(" ")[0])
# stock.saveValues()
# startDate = datetime.datetime(2018, 2, 4, 8) # bull?
# startDate = datetime.datetime(2018, 2, 17, 8) # ?
startDate = datetime.datetime(2018, 3, 6, 8) # bear
endDate = startDate + datetime.timedelta(hours=100)
testStartDate = datetime.datetime(2018, 5, 12, 8)
testEndDate = testStartDate + datetime.timedelta(hours=168)
ratBoi = RatBoi(Stock("BTC-USD"))
ratBoi.train(startDate, endDate, iterations=3, learningRate=0.9)
# ratBoi.test(testStartDate, testEndDate)
ratBoi.simulate(testStartDate, testEndDate, 0)
return
def loadValsForSkewKurtosis(self):
testStock = Stock('TEST', 'testValues.txt')
#3 days, 32 mins span
testStock.record = {
"2019-03-18": {
"06:00:00": (61, 0),
"06:01:00": (61, 0),
"06:02:00": (61, 0),
"06:03:00": (61, 0),
"06:04:00": (61, 0),
"06:05:00": (64, 0),
"06:06:00": (64, 0),
"06:07:00": (64, 0),
"06:08:00": (64, 0),
"06:09:00": (64, 0),
"06:10:00": (64, 0),
"06:11:00": (64, 0),
"06:12:00": (64, 0),
"06:13:00": (64, 0),
"06:14:00": (64, 0),
"06:15:00": (64, 0),
"06:16:00": (64, 0),
"06:17:00": (64, 0),
"06:18:00": (64, 0),
"06:19:00": (64, 0),
"06:20:00": (64, 0),
"06:21:00": (64, 0),
"06:22:00": (64, 0),
"06:23:00": (67, 0),
"06:24:00": (67, 0),
"06:25:00": (67, 0),
"06:26:00": (67, 0),
"06:27:00": (67, 0),
"06:28:00": (67, 0),
"06:29:00": (67, 0),
"06:30:00": (67, 0),
"06:31:00": (67, 0),
"06:32:00": (67, 0),
"06:33:00": (67, 0),
"06:34:00": (67, 0),
"06:35:00": (67, 0),
"06:36:00": (67, 0),
"06:37:00": (67, 0),
"06:38:00": (67, 0),
"06:39:00": (67, 0),
"06:40:00": (67, 0),
"06:41:00": (67, 0),
"06:42:00": (67, 0),
"06:43:00": (67, 0),
"06:44:00": (67, 0),
"06:45:00": (67, 0),
"06:46:00": (67, 0),
"06:47:00": (67, 0),
"06:48:00": (67, 0),
"06:49:00": (67, 0),
"06:50:00": (67, 0),
"06:51:00": (67, 0),
"06:52:00": (67, 0),
"06:53:00": (67, 0),
"06:54:00": (67, 0),
"06:55:00": (67, 0),
"06:56:00": (67, 0),
"06:57:00": (67, 0),
"06:58:00": (67, 0),
"06:59:00": (67, 0),
},
"2019-03-19": {
"06:15:00": (73, 0),
"06:16:00": (73, 0),
"06:17:00": (73, 0),
"06:18:00": (73, 0),
"06:19:00": (73, 0),
"06:20:00": (73, 0),
"06:21:00": (73, 0),
"06:22:00": (73, 0)
},
"2019-03-21": {
"06:01:00": (67, 0),
"06:02:00": (67, 0),
"06:03:00": (67, 0),
"06:04:00": (67, 0),
"06:05:00": (67, 0),
"06:06:00": (70, 0),
"06:07:00": (70, 0),
"06:08:00": (70, 0),
"06:09:00": (70, 0),
"06:10:00": (70, 0),
"06:11:00": (70, 0),
"06:12:00": (70, 0),
"06:13:00": (70, 0),
"06:14:00": (70, 0),
"06:15:00": (70, 0),
"06:16:00": (70, 0),
"06:17:00": (70, 0),
"06:18:00": (70, 0),
"06:19:00": (70, 0),
"06:20:00": (70, 0),
"06:21:00": (70, 0),
"06:22:00": (70, 0),
"06:23:00": (70, 0),
"06:24:00": (70, 0),
"06:25:00": (70, 0),
"06:26:00": (70, 0),
"06:27:00": (70, 0),
"06:28:00": (70, 0),
"06:29:00": (70, 0),
"06:30:00": (70, 0),
"06:31:00": (70, 0),
"06:32:00": (70, 0)
}
}
return testStock
d_bar = 5
interest_rate = 0.02
init_holding = 1
initialcash = 20000
theta = 75
gene_length = 64
risk_coef = 0.5
num_strategies = 80
max_stockprice = 200
min_stockprice = 0.01
min_excess = 0.005
eta = 0.005
specialist_iterations = 10
the_market = Stock(num_shares, init_price, dividend_startvalue, rho, noise_sd,
d_bar, interest_rate)
agents = [
Agent(init_holding, initialcash, num_strategies, theta, gene_length,
risk_coef, num_shares, init_price, dividend_startvalue, rho,
noise_sd, d_bar, interest_rate) for _ in range(100)
]
the_specialist = Specialist(max_stockprice, min_stockprice, min_excess, eta,
specialist_iterations, num_shares, init_price,
dividend_startvalue, rho, noise_sd, d_bar,
interest_rate)
for t in range(1000):
the_market.advance_arprocess()
the_specialist.clear_market(agents)
ma = the_market.calculate_ma(query='dividend', period=50)
def TestCaseWithFit():
from AprFit import AprFit
from winsound import Beep
# Pull the stock data into memory.
stocks = []
for symbol in symbols:
stock = Stock.ShyRetrieve(symbol=symbol, minDate=(datetime.datetime.now() - datetime.timedelta(days=5)))
if len(stock._history) > 0:
stocks.append(stock)
else:
print(f"{stock.symbol} no history")
Beep(300, 500) # Let the user know it's done importing
today = datetime.datetime.now()
# In a loop so data stays in memory if the user wants to change parameters.
while True:
yearsToConsider = AskYears()
if yearsToConsider == "break":
break
scores = []
startDate = today - datetime.timedelta(seconds = 366 * yearsToConsider * 60 * 60 * 24)
for stock in stocks:
if stock.history[0].date > startDate:
continue
# Only plot stocks which have enough history.
try:
apr_fit = stock.get_apr_fit(years=yearsToConsider)
except Exception as ex:
# Looks like I need to do some sanitizing of the data from yFinance. BEP hits a math domain error here and previously gave strange results in other tests.
print(f"Failed to fit curve onto data from {stock.symbol}: {ex}")
continue
# Determine how much this stock is currently overvalued/undervalued
total_dividends_since_t_0 = 0.
index = len(stock.history) - 1
while (stock.history[index].date - today).total_seconds() >= apr_fit.t_0 * 31557600.:
total_dividends_since_t_0 += stock.history[index].dividend
if index == 0:
break
index -= 1
undervalue = apr_fit.y_0 * (1 + apr_fit.rate) ** (0. - apr_fit.t_0) - (stock.history[-1].price + total_dividends_since_t_0)
undervalue /= stock.history[-1].price + total_dividends_since_t_0
N_STDEVS = 2. # Number of standard deviations to use for fitness
if apr_fit.stdev * N_STDEVS > 1.:
score = Score(stock, 0.)
else:
score = Score(stock, (apr_fit.rate + undervalue / 4.) * (1. - N_STDEVS * apr_fit.stdev) + 1. / stock.pe_ratio)
# stdev applies as both uncertainty in the fit and as potential loss
if math.isnan(score.score):
print(f"Failed to calculate score for {stock.symbol}")
continue
# Boost the investments we want to bias towards
if stock.symbol in boost:
score.score *= 1.5
# Nerf the stocks we don't like
if stock.symbol in nerf:
score.score *= 0.5
# Attach metadata to the score so we can print it.
score.rate = 100. * apr_fit.rate
score.undervalue = 100. * undervalue
score.uncertainty = 100. * apr_fit.stdev
scores.append(score)
scores.sort(key=lambda x:x.score)
scores.reverse()
# Recommend how much would have been good to allocated to each
number_of_recommendations = 0
sum = 0
for candidate in scores:
if math.isnan(candidate.score):
continue
if candidate.score <= 0.:
continue
sum += candidate.score
if candidate.score / sum < 0.05:
sum -= candidate.score
break
number_of_recommendations += 1
# Don't recommend >25% in any one asset
LIMIT = 0.25
if scores[0].score > sum * LIMIT:
redistribution = (scores[0].score - sum * LIMIT) / (number_of_recommendations - 1)
for score in scores:
score.score += redistribution
if scores[1].score > sum * LIMIT:
redistribution = (scores[1].score - sum * LIMIT) / (number_of_recommendations - 2)
for score in scores:
score.score += redistribution
scores[1].score = sum * LIMIT
scores[0].score = sum * LIMIT
print("Recommended distribution:")
for i in range(number_of_recommendations):
stock = scores[i].stock
recommended_percent = 100 * scores[i].score / sum
print(f"{recommended_percent:.2f}% in {stock.symbol} Score: {scores[i].score:.2f} P/E: {stock.pe_ratio:.2f} <APR>: {scores[i].rate:.2f} undervalued by {scores[i].undervalue:.2f}% uncertainty: {scores[i].uncertainty:.2f}%")
if plot:
stock.get_apr_fit(years=yearsToConsider, plot=True)
Beep(300, 500)
def OriginalTestCase():
# Import the modules now so it will crash before doing work if a module is missing.
import matplotlib.pyplot as plt
from dateutil.parser import parse
# Pull the stock data into memory.
stocks = []
for symbol in symbols:
stock = Stock.ShyRetrieve(symbol=symbol, minDate=(datetime.datetime.now() - datetime.timedelta(days=5)))
if len(stock._history) > 0:
stocks.append(stock)
today = datetime.datetime.now()
# Let the user know import is complete
import winsound
winsound.Beep(600, 100)
# In a loop so data stays in memory if the user wants to change parameters.
while True:
yearsToConsider = AskYears()
minimumYield = AskMinYield()
if yearsToConsider == "break":
break
if minimumYield == "break":
break
startDate = today - datetime.timedelta(days = 365 * yearsToConsider)
if plot:
print("Plotting...")
fig, subplots = plt.subplots(nrows=2, ncols=1)
dividend_plot = subplots[0]
#growth_plot = subplots[1]
total_yield_plot = subplots[1]
dividend_plot.set_ylabel("Dividend (%/yr)")
dividend_plot.set_xlim(left=startDate, right=today)
max_dividend = 0.
min_dividend = 0.
#growth_plot.set_ylabel("Growth Since (%/yr inflation adjusted)")
#growth_plot.set_xlim(left=startDate, right=today)
max_growth = 0.
min_growth = 0.
total_yield_plot.set_ylabel("Total Yield (%/yr)")
total_yield_plot.set_xlim(left=startDate, right=today)
max_total = 0.
min_total = 0.
annual_yields = []
scores = []
for stock in stocks:
if stock.history[0].date > startDate:
continue
# Only plot stocks which have enough history.
filteredGrowth, growthUncertainty = stock.GrowthAPRWithUncertainty(yearsToConsider)
dividendYield = stock.AverageDividendPercent(yearsToConsider)
dividendUncertainty = stock.DividendPercentUncertainty(yearsToConsider)
if dividendUncertainty > dividendYield:
dividendUncertainty = dividendYield # Never apply a penalty for dividends
annualYield = dividendYield + filteredGrowth
if annualYield < minimumYield:
continue
# Only plot stocks with combined margin better than 10% per year over the past 10 years.
print(f"{stock.symbol} {annualYield:.1f} ± {growthUncertainty + dividendUncertainty:.1f}% average annual yield over the past {yearsToConsider} years.")
print(f"{filteredGrowth:.2f}% from growth and {stock.AverageDividendPercent(yearsToConsider):.2f}% from dividends.")
annual_yields.append((stock.symbol, annualYield))
score = annualYield - 0.431 * (growthUncertainty + dividendUncertainty) # 33rd percentile
scores.append((stock.symbol, score))
if plot:
dates = []
relGrowth = []
divYieldPercent = []
latest = stock.history[-1]
for snapshot in stock.history:
if snapshot.date < startDate:
continue
if snapshot.price == 0.0:
continue
if (latest.date - snapshot.date).days == 0.0:
continue
annual_growth = 100. * (latest.price / snapshot.price - 1.) * 1.02 ** ((snapshot.date - latest.date).days / 365.25) / \
((latest.date - snapshot.date).days / 365.25)
if annual_growth > 1. * (latest.date - snapshot.date).days:
annual_growth = 1. * (latest.date - snapshot.date).days
if annual_growth < -1. * (latest.date - snapshot.date).days:
annual_growth = -1. * (latest.date - snapshot.date).days
dates.append(snapshot.date)
relGrowth.append(annual_growth)
divYieldPercent.append(100. * snapshot.annualDividend / snapshot.price)
if relGrowth[-1] > max_growth:
max_growth = relGrowth[-1]
elif relGrowth[-1] < min_growth:
min_growth = relGrowth[-1]
if divYieldPercent[-1] > 100.:
print(f"{dates[-1]}: {stock.symbol} has {divYieldPercent[-1]}% dividend?")
print("You may not want to trust this data...")
if divYieldPercent[-1] > max_dividend:
max_dividend = divYieldPercent[-1]
elif divYieldPercent[-1] < min_dividend:
min_dividend = divYieldPercent[-1]
dividend_plot.plot(dates, divYieldPercent, label=stock.symbol)
#growth_plot.plot(dates, relGrowth, label=stock.symbol)
annualYield = []
yDates = []
for index in range(200, len(stock.history)):
snapshot = stock.history[index]
if snapshot.date < startDate:
continue
priceLastYear = stock.history[index - 200].price
if priceLastYear == 0.0:
continue
yDates.append(snapshot.date)
annualYield.append(100. * ((snapshot.price - priceLastYear) + snapshot.annualDividend) / priceLastYear)
if annualYield[-1] > max_total:
max_total = annualYield[-1]
elif annualYield[-1] < min_total:
min_total = annualYield[-1]
if plot:
total_yield_plot.plot(yDates, annualYield, label=stock.symbol)
annual_yields.sort(key=lambda x:x[1])
annual_yields.reverse()
scores.sort(key=lambda x:x[1])
scores.reverse()
# Recommend how much would have been good to allocated to each
number_of_recommendations = 0
sum = 0
for candidate in scores:
if not math.isnan(candidate[1]):
sum += candidate[1]
if sum > 0.:
if candidate[1] / sum < 0.05:
sum -= candidate[1]
break
if sum < 0.:
sum -= candidate[1]
break
number_of_recommendations += 1
print("Recommended distribution:")
recommendations = []
for i in range(number_of_recommendations):
symbol = scores[i][0]
recommended_percent = 100 * scores[i][1] / sum
recommendations.append((symbol, recommended_percent))
print(f"{recommended_percent:.2f}% in {symbol} Score: {scores[i][1]:.2f}")
winsound.Beep(300, 500)
if plot:
fig.tight_layout()
fig.autofmt_xdate()
#growth_plot.plot([parse("1/1/2000"), today], [0, 0], label='Zero')
total_yield_plot.plot([parse("1/1/2000"), today], [0, 0], label='Zero')
dividend_plot.set_ylim(bottom=min_dividend, top=max_dividend)
#growth_plot.set_ylim(bottom=min_growth, top=max_growth)
total_yield_plot.set_ylim(bottom=min_total, top=max_total)
dividend_plot.tick_params(labelleft=True, left=True, right=True, bottom=True)
#growth_plot.tick_params(labelleft=True, left=True, right=True, bottom=True)
total_yield_plot.tick_params(labelleft=True, left=True, right=True, bottom=True, labelbottom=True)
dividend_plot.legend()
#growth_plot.legend()
#total_yield_plot.legend()
plt.show()
def setUp(self):
self.TEA = Stock("TEA", "Common", last_dividend=0, par_value=100)
self.POP = Stock("POP", "Common", last_dividend=8, par_value=100)
self.ALE = Stock("ALE", "Common", last_dividend=23, par_value=60)
self.GIN_PRE = Stock("GIN", "Preferred", last_dividend=8, par_value=100, fixed_dividend=0.02)
self.GIN_COM = Stock("GIN", "Common", last_dividend=8, par_value=100)
from Stocks import Stock
ALL_STOCKS = {
'amazon': Stock('AMZN', 'amazonValues.txt'),
'walmart': Stock('WMT', 'walmartValues.txt'),
'generalElectric': Stock('GE', 'generalElectricValues.txt'),
'cocaCola': Stock('KO', 'cocaColaValues.txt'),
'google': Stock('GOOGL', 'googleValues.txt'),
'microsoft': Stock('MSFT', 'microsoftValues.txt'),
'tesla': Stock('TSLA', 'teslaValues.txt'),
'apple': Stock('AAPL', 'appleValues.txt'),
'facebook': Stock('FB', 'facebookValues.txt'),
}